Emergency rooms serve as the point of entry for nearly half of all hospital admissions in the United States. About 120 million Americans visit an emergency room each year—a number that continues to rise, even though there are 22 percent fewer emergency departments today than there were two decades ago. These departments, as a result, are in a state of overcapacity roughly 50 percent of the time. Patients wait about an hour, on average, before seeing a physician, and sometimes several hours. These people may be in serious discomfort and, in extreme cases, some may even die before a doctor gets to them.
A cost-effective system that reduces waiting times would surely alleviate this growing problem. Assistant Professor of Public Policy Soroush Saghafian, whose research focuses on hospital efficiency, took up this challenge with four colleagues in an August 2017 faculty research working paper, “Workload Management in Telemedical Physician Triage and Other Knowledge-Based Service Systems.” Saghafian believes that help on this front could come from his field of specialty, operations research, which relies on mathematical modeling, statistics, and other analytic techniques to assist in decision-making.
In their paper, Saghafian and his coauthors assessed the potential of telemedicine, which draws on the services of remotely based physicians, to ease emergency room congestion. This could be an economical option for hospitals, which may not be able to afford hiring more doctors and nurses or adding more beds.
The model the researchers devised had a hierarchical structure. At the first level were triage nurses, who evaluated patients and had to determine, as quickly as possible, whether they should be referred to more knowledgeable telemedical physicians, occupying the second tier. These physicians, in turn, used the internet to communicate with front-line nurses via the internet, monitor a patient’s vital signs using advanced technology, order more tests when warranted, and perhaps expedite admittance to the hospital.
Saghafian and his collaborators employed queuing theory, the mathematical study of waiting lines, to optimize a system that would confer the most benefits at the least cost, both to the patients and the hospital. Their 2017 paper, that will appear later in the journal Management Science, concluded that telemedicine could be a useful resource for hospitals, especially during periods of overcrowding. “Because triage classifications are used for prioritization, they need to be viewed as relative rather than absolute ratings,” the authors noted. “This is in sharp contrast with the prevailing belief that triage classification should depend only on the medical conditions of a patient.”
The authors also found that telemedical physicians can make better decisions, in principle, when triage nurses share their judgments with them. However, this input was only valuable when nurses were consistent in their classifications, even though the doctors were, by virtue of their education and experience, more accurate in their diagnoses. Therefore, in the training of triage nurses, consistency should be emphasized over accuracy. “If the nurses are not consistent, their appraisals are of little use,” explains Saghafian. “But if their assessments are consistent, even if not wholly accurate, doctors can put that information into perspective and use it.”
As a next step, Saghafian is trying to get the word out and encourage more emergency departments to adopt “telemedical physician triage.” In current research, he is exploring broader uses of this approach in hospitals outside of emergency rooms. Saghafian is also looking into the advantages of making public reporting of hospitals more transparent, “so that patients can be better equipped to choose the hospital they want, just as they might do when choosing which car to buy,” he says. “If I knew, for instance, that 10 percent of the patients with a condition similar to mine died in one hospital, whereas only 1 percent died in another hospital, I might be inclined to go to the second establishment.”